It’s high time for military technology specifiers and their suppliers in the aerospace and defense industries to come up with a new electronics packaging standard to complement-and sometimes replace-the venerable air transport rack (ATR) described in ARINC 404A.
To be sure, there’s nothing wrong with the ATR that cries out for replacement, for it continues to represent the typical avionics subsystem “box” that has been in reliable use since the 1970s to contain standard electronic printed-circuit cards.
Yet, while the ATR line-replaceable unit (LRU) undoubtedly will continue in useful service for years to come, several design trends are converging to form a pressing need for something more.
Perhaps the biggest of those trends are miniaturization in electronic subsystems for new platforms like unmanned aerial vehicles (UAVs) and wheeled combat vehicles, as well as the evolution of high-speed serial-fabric networks.
The ATR and its subsystem-box cousins are familiar fixtures in avionics and other embedded electronics architectures. Coming in quarter, half, full, and other ATR sizes, they typically have parallel backplane databuses such as VME and CompactPCI into which systems designers plug circuit cards. ATR boxes themselves are connected on wired or optical-fiber serial networks such as MIL-STD-1553, ARINC 429, Fibre Channel, RapidIO, StarFabric, or Gigabit Ethernet.
When designers talk about electronics upgrades that are “form-, fit-, and function-compatible” with legacy architectures, essentially they mean that new boxes, circuit cards, and components fit into the same spaces as the old ATR boxes.
While that approach has its advantages in predictability and standardization, it can also lead to kludgy solutions in which-because of the ever-shrinking nature of electronic components-the new ATR boxes contain more empty air than circuit cards. After all, the upgraded solution has to fit in the old ATR-size hole, whether or not the new components are smaller than the ones they replace-which they usually are.
A new approach could change all that.
Think blade servers. These essentially are single-board computers with support electronics packaged in thin metal containers stacked in a rack and connected on high-speed serial-data networks. No backplane databus is necessary.
The rack needs only to be as high as the number of blades necessary to do the job at hand. Place more functionality onto each blade, and the designer can shrink the size of the architecture by using fewer blades, or pack more functionality into the same space. Get the picture?
We could package these units in whatever size makes sense, as long as everyone concerned in industry signs off on the standard. An approach like this could accommodate standard circuit card sizes such as 3U, 6U, and 9U VME and Compact PCI, while enabling these cards to work in more functional and flexible embedded architectures.
We could call these units line-replaceable modules, or LRMs-a parallel to the LRUs that the ATR box represents. The LRM package itself could double as protection for the card inside from shock, vibration, temperature extremes, and other environmental hazards to reduce the need for circuit card conformal coatings and stiffeners. In addition, the LRM standard could accommodate convection, conduction, or flow-through cooling at the LRM level, rather than at the card level.
Now think about an avionics upgrade involving these LRMs that does not limit designers to ATR-box-size spaces. Instead, designers could fill every nook and cranny-weight and balance permitting-with LRM functionality.
Think also about new electronic architectures on UAVs, combat land vehicles, wearable computers, mobile communications systems, and other applications where size and weight are crucial. Designers could put in just the functionality they need-no more, no less-with a real promise for vastly increased functionality with successive generations of upgrades, and involving virtually no size penalties.
Not only would such an approach facilitate systems upgrades and offer new perspectives on electronic architectures, but it also could enable systems integrators to think about designing by function, rather than by technology. It would be simple, elegant, and make a lot of sense.
A standard like this offers the potential to augment the ATR box with something more flexible. It would take off some of the pressure to ruggedize from circuit-card designers and let them focus instead on functionality. It would also offer a flexible electronics architecture that is just as applicable to UAVs as it is to land vehicles, ships, and legacy aircraft upgrades.
Committees of VITA, PICMG, and other standards organizations are scheduling important meetings for this spring. I hope an idea like this gets on their agendas.
